Antioxidant and toxicological potential of the Golden trumpet hydroalcoholic stem bark extract

Handroanthus chrysotrichus is a tree of the Bignoniaceae family known as golden trumpet that is distributed throughout Northeast, Southeast and South Brazil. Its flowers, stem and bark are used for medicinal purposes in the treatment of cardiovascular and immune system diseases. This study aims to evaluate the phytochemical profile, biological activity spectrum, antioxidant capacity and toxicological potential of H. chrysotrichus stem bark extract. Hydroethanolic extract was obtained by percolation and lyophilized. Compounds present in the extract were analyzed by colorimetric methods and by GC-MS. Evaluation of the biological activity spectrum was performed in silico. Antioxidant power was determined by investigation of total antioxidant capacity, iron chelating capacity, DPPH and ABTS assays, and deoxyribose degradation test. The ability to inhibit Fe induced lipoperoxidation was evaluated in mouse brains and livers. Nauplii of Artemia salina were used to evaluate the median lethal dose. Toxicity was assessed by computer simulation, and in vitro in human lymphocytes. As a result, colorimetric methods suggest high levels of polyphenols and GC-MS data indicated the occurrence of α-curcumene, β-bisabolene, 4(4-methylphenyl) pentanal, pentanoic acid and isoamyl acetate. Computer simulations have pointed biological activities that are in accordance with their traditional use. The H. chrysotrichus stem bark extract exhibited antioxidant activity in several assays and was effective in protecting mouse brains and livers from Fe induced lipoperoxidation. H. chrysotrichus stem bark extract showed medium toxicity in A. salina with potential presence of bioactive compounds. In general, the compounds showed low probability of toxicity in silico predictions. There was no cytotoxicity and genotoxicity in human lymphocyte evaluation. The results indicate that H. chrysotrichus stem bark extract has compounds with biological activity spectrum and low toxicological potential. It also shows antioxidant capacity and protective action against lipid peroxidation. The data presented support the medicinal use of golden trumpet and point to it as a promising extract for in vivo evaluations.

evaluated in mouse brains and livers. Nauplii of Artemia salina were used to evaluate the median lethal dose. Toxicity was assessed by computer simulation, and in vitro in human lymphocytes. As a result, colorimetric methods suggest high levels of polyphenols and GC-MS data indicated the occurrence of α-curcumene, β-bisabolene, 4-(4-methylphenyl) pentanal, pentanoic acid and isoamyl acetate.
Computer simulations have pointed biological activities that are in accordance with their traditional use.
The H. chrysotrichus stem bark extract exhibited antioxidant activity in several assays and was effective in protecting mouse brains and livers from Fe + induced lipoperoxidation. H. chrysotrichus stem bark extract showed medium toxicity in A. salina with potential presence of bioactive compounds. In general, the compounds showed low probability of toxicity in silico predictions. There was no cytotoxicity and genotoxicity in human lymphocyte evaluation. The results indicate that H. chrysotrichus stem bark extract has compounds with biological activity spectrum and low toxicological potential. It also shows antioxidant capacity and protective action against lipid peroxidation. The data presented support the medicinal use of golden trumpet and point to it as a promising extract for in vivo evaluations.

Introduction
The family Bignoniaceae is the major group of Angiosperms plants and has been used for its beneficial health properties. Several parts of Bignoniaceae species as leaves, fruits, roots, sap, flowers and bark are traditionally used for treatment of diabetes, high blood pressure, asthma, cancer, uterine infection and others diseases. In all these uses, different extractive methods are employed including decoction, maceration, infusion, poultice, syrup and tincture (Bolson et al., 2015;Ribeiro et al., 2017).
In this way, the scientific literature reports many therapeutics properties for Handroanthus. For example, species H. impetiginosus showed analgesic, anti-inflammatory and antiulcerogenic effects in animal models Twardowschy et al., 2008). H. impetiginosus has presented antibacterial and antioxidant activity (Park et al., 2003;Park et al., 2006). H. chrysanthus demonstrated immunostimulant activity (Perez et al., 2004) and T. aurea decreases inflammatory, myotoxic and hemorrhagic activities induced by snake venom (Reis et al., 2014). These studies point confirmations that family Bignoniaceae is a promising group with pharmacological activities.
Thus, considering the H. chrysotrichus empirical use, this study aimed: a) to quantify the total polyphenols and flavonoids contents; b) to identify the phytochemical major compounds in bark extract; c) to evaluate the biological activity spectrum of identified major compounds; d) to determine the antioxidant capacity of extract; d) to verify the median lethal dose, cytotoxicity and genotoxicity for crude extract and major compounds.

Phytochemical Evaluation
Total phenolic and flavonoid content of HCBE was measured according Nurmi et al. (1996) and Choi et al. (2002), respectively. A standard curve of gallic acid and quercetin was used to determine the polyphenols and flavonoids equivalents content. GC-MS analyses were performed according Soares et al. (2017), with some modifications. Briefly, the oven temperature program was as follows: initial oven temperature was held at 50ºC for 5 min, and then increased to 150ºC at a rate of 10ºC. min -1 for 10 min, after was increased to 200ºC at a rate of 10ºC. min -1 and detained for 01 min and finally increased to 280ºC at a rate of 10ºC.
min -1 and held for 10 min. Ion source and transfer line temperatures were 280ºC. Compounds were identified by comparing mass spectra with data from the NIST library that are available in the instrument.

In silico Evaluations of GC-MS Identified Compounds
Biological activity spectrum The computational Pass (Prediction of Activity Spectra for Substances, available in http://www.pharmaexpert.ru/PASSonline/predict.php) was applied on five major compounds identified in HCBE (Drwal & Griffith, 2013) to search potential pharmacological actions.
Results were expressed in percentage of probable activity (Pa) and probable inactivity (Pi). Pa and Pi values vary from 0.000 to 1.000 and it was considered significant activity with Pa > Pi and Pa > 0.700.

HCBE Antioxidant Analyses
Total antioxidant capacity (TAC), DPPH • and ABTS •+ scavenger assay TAC of HCBE was measured through spectrophotometric method proposed by Prieto et al. (1999). HCBE activity against DPPH • and ABTS •+ radicals were performed in accordance with Choi et al. (2002) and Re et al. (1999), respectively. Results were expressed as inhibitory concentration of 50% (IC50), based on percentage of radical inhibition in relation to the control without extract.

Ferric reducing antioxidant power (FRAP) assay
FRAP assay was measured for HCBE by spectrophotometric method (Benzie & Strain, 1996) and determined by plotting in a standard curve produced by the addition of ferrous sulfate to the FRAP reagent. Results were expressed as IC50, based on the extract ability to reduce Fe 3+ to Fe 2+ .

Thiobarbituric acid reactive substances (TBA-RS) assay
TBA-RS was measured according Ohkawa et al. (1979) and used as an index of 2deoxy-D-ribose degradation, and lipid peroxidation marker in brain and liver of Swiss albino mice, as described below. a) Deoxyribose degradation assay The deoxyribose degradation assay was performed according Puntel et al. (2005) with modifications. Briefly, the reaction medium was prepared containing: HCBE (concentrations 0 -1000 μg. mL -1 ), 2mM deoxyribose, 0.05mM and FeSO4. After incubation at 37ºC for 60 min, the reaction was stopped with 2% of trichloroacetic acid (TCA). Color reaction started with addition of 0.4mL of TBA and allowed to incubate for 30 min at 100ºC. Standard curves of malondialdehyde (MDA) were performed to determine the MDA generated by the deoxyribose degradation, and the values were expressed as percentage of control (blank). b) Analyses with brain and liver of Swiss albino mice These biological tissues were donated from another projects where animals were maintained and used in accordance with guidelines of the Committee on Care and Use Research, Society and Development, v. 9, n. 4, e122942936, 2020 (CC BY 4. 30-35g) were euthanized by decapitation and the brain and liver were quickly removed, homogenized in NaCl (150 mM) and kept in ice. After homogenization, samples were centrifuged at 2000g at 4°C for 10 min to yield a low speed supernatant fraction (S1).
The obtained S1 was mixed with ferrous sulfate (FeSO4) in concentrations of 0.01mM, with or without HCBE (concentrations 0 -1000 μg. mL -1 ). A standard curve of MDA was constructed to determine TBA-RS content. Results are expressed as IC50, based on the extract ability to inhibit lipid peroxidation of tissue, corrected by mg of tissue.
Briefly, A. salina cysts were induced to hatch in aerating solution (saline water 3%) for 24 h.
The nauplii were collected and transferred individually to a 96-well plate containing different HCBE concentrations (50, 100, 300, 600 and 900 µg. mL -1 ) or control saline solution. The mortality was analyzed after 24 h. Assays were performed thrice in triplicate, with n=180 nauplii in each assay. The median lethal dose (LD50) was the required concentration to kill 50% of nauplii. It was considered LD50 < 1000 µg. mL -1 as toxic and a possible presence of bioactive compounds. Genotoxicity was evaluated in PBMC through comet assay according to Singh et al. (1988). After mounting slides and performing electrophoresis, the lengths of tails were measured under fluorescence microscopy after adding ethidium bromide on the slides.
One hundred cells from each of the three replicate slides were analyzed. Cells were visually scored according to tail length and receive scores from 0 (no migration) to 4 (maximal migration). Analyses was based on the reading damage index, which represented the sum of cells identified in each class multiplied by the class.

Statistical Analyses
Data were analyzed by two-way ANOVA followed by Tukey's multiple comparisons Test. Cell viability assays were interpreted by one-way ANOVA followed by Bonferroni's post hoc Test. Genotoxicity were analyzed by nonparametric Kruskal-Wallis with Dunn's Test.
Values of p < 0.05 were considered significant. Data are presented as mean and standard deviation (SD). IC50 and LD50 determinations were performed using a logarithmic regression curve.

In silico Evaluations HCBE Major Compounds
The possible biological activity spectrum related to the five HCBE major compounds was evaluated by online platform PASS (Table 2). We selected the predicted properties with values of Pa > 0.7. Our data indicate that α-curcumene, β-bisabolene, 4-(4-methylphenyl) pentanal, pentanoic acid and isoamyl acetate, respectively, presented properties in common as mucobembranous protector and antieczematic. Others frequently properties in compounds are its carminative and fibrinolytic potential.
Probable activity; Pi = Probable inactivity. Pa > 0.700 = probable activity greater than 70%. The PASS prediction results were interpreted and used as follows: (i) only activities with Pa > Pi are considered as possible for a particular compound; (ii) if Pa > 0.7, the chance to find the activity experimentally is high.  , Society and Development, v. 9, n. 4, e122942936, 2020(CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v9i4.2936 The toxic risk predictions performed (mutagenic, carcinogenic, cardiotoxic, hepatotoxic, skin sensitization and reproductive toxicity) for α-curcumene, β-bisabolene, 4-(4methylphenyl) pentanal, pentanoic acid and isoamyl acetate were assessed by five online platforms and the results are present in Table 3. In general, the compounds displayed low toxicity probability. However, β-bisabolene and α-curcumene showed medium and high carcinogenic potential. For the α-curcumene, 4-(4-methylphenyl) pentanal and isoamyl acetate, the results showed a skin sensitization probability on two of the platforms employed. In the other evaluated parameters, isoamyl acetate and pentanoic acid demonstrated high risks to reproductive system. Moreover, theoretical toxicity of compounds in the others tools employed suggesting a low toxicity risk of the compounds.  Values are expressed as mean ± SD (n=3).

Phytoconstituents Main predicted properties by PASS online
Analyses with tissues of Swiss albino mice exposed to Fe ++ showed the protective capacity of HCBE from 100 and 50 µg. mL -1 for brain and liver, respectively (Figure 3). In this context, the extract decreased TBA-RS levels for tissue reported.

Discussion
This study aimed to evaluate the toxicological and antioxidant potential of HCBE in silico and in vitro. The extract showed five major compounds identified with pharmacological potential and low toxicity risks predictions in silico. In addition, HCBE presented antioxidant activity, protected brain and liver of mice against oxidative damage and no present cell toxicity in vitro. These data may justify its use in folk medicine, and it proves to be a promising extract for in vivo testing.
Initially, A. salina assays resulted medium toxicity. It indicate bioactive compounds presence, besides a positive correlation with cytotoxicity in cancer cell lines (McLaughlin, 1991). In agreement GC/MS analyses suggested the presence of substances as β-bisabolene which it was reported tumor-specific pro-apoptotic properties (Yeo et al., 2016). Although tests of α-curcumene in human colon cancer cell line did not inhibit cell proliferation (Tyagi et al., 2015).
Major compounds found α-curcumene and β-bisabolene are terpenes also presents in bark of H. heptaphyllus (Garcez et al., 2007) and act among other things against insects protecting the plant (Júnior, 2003). These insecticidal functions promoted the mortality of H. impetiginosus bark also presented larvicidal activity in three mosquitos species and H. serratifolius steam had antileishmanial activity (Costa et al., 2017;Kim et al., 2013).
Other compounds are ester isoamyl acetate, well known in the food industry (Torres et al., 2010); and the short-chain fatty acid (SCFA) pentanoic acid, a superior class of cellulosic biofuels. This SCFA may be produced by intestinal bacteria (Topping, 1996) and induces thymic stromal lymphopoietin production (Mizuno et al., 2017). Thymic stromal lymphopoietin is an IL-7-like cytokine mainly derived from fibroblasts and epithelial cells and is related to allergic inflammation (Liu, 2006;Sims et al., 2000). Nevertheless, 4-(4-Methylphenyl) pentanal is not target of many studies.
Throughout chromatography analyzes β-lapachone and lapachol was expected in extract studied, since they are cited in literature as major compound in H. chrysotricus (Grazziotin et al., 1992), H. impetiginosus (Castellanos et al., 2009;Ferreira et al., 1989) and others Bignoniaceae plants. In addition, lapachol is soluble in ethanol (Kiage-mokua et al., 2012), being consistent with method here applied, but this compound was not detected by High Performance Liquid Chromatography (Supplementary material). Thus, it seems to be absent in the H. chrysotricus bark analyzed. These substantial differences between specimen may be due to environmental conditions, because lapachol presence may vary in the same specie according rainfall, temperature and the soil content (Wright & Setzer, 2013).
The biological activity spectrum in silico of isolated compounds (Table 2) showed compatible action with popular practice. In this line, Salgueiro et al. (2018) suggest that in silico investigations can be used as a complementary route after the ethnopharmacological research.
Indeed, in silico studies may be useful to point out several possibilities for conducting in vitro and in vivo studies (Salgueiro et al., 2018). Brazilian ethnopharmacological investigations indicated therapeutic use of genus Handroanthus as anti-inflammatory and for cancer treatment, back pain, stomachache, gastritis, ulcer and allergy (Bieski et al., 2015;Ribeiro et al., 2017). According to Bolson et al. (2015) H. chrysotrichus also may be used for cardiovascular system diseases, immunological and poisoning. Moreover, other in silico analyses were made, and the results showed low toxicity probability in most of the analyzed criteria (Table 3). Nevertheless, care should be taken with carcinogenic potential and topical use of the extract, due skin sensitization of some compounds.
The biological activity spectrum in silico no showed, but the HCBE had significant antioxidant activity (Table 4). These important effect may reduce oxidative damage induced by several diseases (Andréia Caroline Fernandes Salgueiro et al., 2016). The absence of antioxidant activities in in silico investigations can be explained by synergism among different molecules in a crude plant extract (Salgueiro et al., 2018). In the context of computational evaluations, the interaction between compounds cannot be evaluated (Salgueiro et al., 2018).
The extract exhibited better activities in FRAP, TAC and ABTS •+ scavenger. Antioxidant analyses as ferric reducing antioxidant power assay, FRAP, propose the affinity of HCBE with Fe ions. We know that iron is an essential element for the body and may induce oxidative stress via Fenton reaction (Benzie & Strain, 1996;Halliwell & Gutteridge, 2007). In addition, HCBE decreased Fe + -induced damage to brain and liver tissues of mice (Figure 3). TAC and ABTS •+ are based on the reduction of molybdenum and ABTS radicals, respectively. ABTS •+ assay is a simple method applicable to the extracts antioxidant study, but this radical is not found in biologicals systems (Magalhães et al., 2008;Prieto et al., 1999;Re et al., 1999). Therefore, these antioxidants properties may be linked to polyphenols found in the colorimetric phytochemical analyses. In fact, polyphenols have a high reactivity and ability to stabilize unpaired electron, including to chelate transition metal ions (Vertuani et al., 2004). On the other hand, antioxidant activity against DPPH • showed high IC50, compatible data with bark extract of T. rosea (Jimenez-Gonzalez et al., 2018), suggesting that bark extracts these trees have low efficiency in this test. The extract here analyzed not indicated reaction with nitric oxide (NO • ) (data not show), a pro-inflammatory mediator that plays an important role in the regulation of immune functions, neurotransmission, and vasodilation (Sharma, Al-Omran, & Parvathy, 2007).
Despite knowledge about vegetal extracts are necessary, because shows potential actions and may value the plant, we considered important to perform in vitro assays evaluating the cell responses (Fraga et al., 2020;Lima et al., 2020). There was not changes of viability and genotoxicity in cells exposed to HCBE (Table 5). These data are according to Boriollo et al. (2017), which suggested absence of genotoxicity of the H. impetiginosa bark. In addition, confirm the low toxicity found in in silico analyses.
Therefore, the species studied showed low toxic potential and exhibited a biological activity spectrum, including the antioxidant capacity and the protective action in liver and brain of mice. Literature suggested that its major compounds presents biological act. These data support point the HCBE as a promising extract for in vivo evaluations.
The results also verify high levels of polyphenols and show that H. chrysotrichus bark has antioxidant effect in different systems. Despite the extract had presented toxicity in Artemia satureioides bioassay, there was not cell toxicity. Finally, our tests indicate H. chrysotrichus as a potential agent for in vivo testing, especially for the treatment of diseases in which the bark is used and to ensure safety the medicinal use of this plant.